|
II. Scientific Evidence Regarding Rainforest Ecology and Protection
9. OTHER ECOLOGICAL DISTURBANCES
Soil chemistry and water runoff will impact on rainforest topographically
below a coupe (Camerson and Henderson 1979). Some rainforests
are located on steep slopes where soils in adjacent sclerophyll
forest may be prone to be unstable, resulting in erosion, landslips,
silting, increased runoff, changes in soil chemistry, and leaching
of soil nutrients in adjacent rainforest. In such circumstances,
the Code of Forest Practice limits many of the impacts (see below),
and restrictions on falling trees into buffers means frequently,
buffers on steep slopes are considerably broader than 20 metres.
Nevertheless, hydrology, soil chemistry and water runoff may
impact on rainforest topographically below a coupe. Stream flow
is significantly influenced by forest age (Kuczera 1987). This
relationship can be explained by variation in transpiration (Jayasuriya
et al. 1993). Such dynamics will affect the status of
the unsaturated soil zone and may affect moisture balance in rainforest
stands, particularly over summer. The consequences of such changes
for moisture dependent rainforest epiphyte species are unknown
but may be significant (see above). The drying effect of regrowth
will depend on forest type and the proportion of the catchment
logged.
Typically, harvesting operations affect most measurable soil physical
and chemical parameters such as particulate organic matter, soil
moisture, total pore space, aeration porosity, available water
holding capacity, particle size, bulk density, compressive strength
and shear strength (eg., Jusoff and Majid 1992). Ellis and Graley
(1987) described differences in soil chemical composition and
rates of mineralisation of nitrogen in soils from forests in different
successional stages between eucalypt forest and rainforest. Forest
clearing results in a complex series of changes in soil structural
attributes, organic inputs and processing rates, the impacts of
which are felt primarily during downhill transport of material
during storms (Golladay et al. 1992, Webster et al.
1990). It is very difficult to predict what, if any, important
impact there will be on the ecology of rainforest from changes
in the soil composition and hydrological dynamics of forests upslope.
The response of soil water and soil chemistry to clear falling
will almost certainly depend on many parameters including slope,
soil physical and structural parameters and the climatic conditions
preceding and following the clear falling operation. These issues
are recognised as a potential problem in Victoria and specifications
in the Code of Forest Practices (CFL 1989) have been designed
to deal with them including slope limitations, batters, snig track
bars, stream and drainage line crossing specifications, road maintenance,
and wet weather restrictions.
It is likely that access tracks, roads, or extraction activities,
may impinge on the rainforest, sclerophyll forest ecotone where
it extends beyond 20m from rainforest and on the rainforest itself
in places where it is necessary to cross drainage lines through
rainforest. Such disturbances are limited in extent, though they
are likely to affect the regeneration of rainforest species in
places where they occur (Calais and Kirkpatrick 1983, Horne and
Hickey 1991). Certainly, roading has the potential to increase
stream sediment loads significantly, even when well maintained
(Grayson et al. 1993). The Department should consider
the consolidation of disturbance areas into adjacent stands to
minimise the impact of roading on rainforest, a process recommended
by CNR (1993) and followed by CNR (1994). The intention of this
procedure is in part to minimise the amount of road construction
activity, particularly in cool temperate rainforest stands.
Disturbance up to the buffer boundary results in increased risk
of penetration of rainforest by exotic species and disturbance
specialists. In patchy, disturbed environments, plant dispersal
mechanisms are such that undisturbed stands adjacent to disturbance
are subject to a much higher seed rain from secondary successional
and opportunistic species than they otherwise would be. The extent
of the rain depends on the biology of the species, but may extend
many hundreds of metres beyond the boundary of disturbed/undisturbed
tropical forest habitat (eg., Janzen 1986, Laurance 1991).
Carr et al. (1992, p. 14) listed 18 weed species in warm
temperate rainforest and 7 weed species in cool temperate rainforest
that pose 'very serious' threats to those communities, and noted
that weed prevalence is related to the kind and proximity of disturbance.
In cool temperate rainforest, foxglove (Digitalis purpurea),
blackberry (Rubus fruticosus) and clovers (Trifolium
spp.) are particularly serious potential problems (Cameron 1987)
and there are numerous potential indigenous and non-indigenous
Australian native weed species (McMahon 1987, Carr et al.
1992; see, for example, Neyland and Brown 1994). Secondary rainforest
and sclerophyll opportunist species including Prostanthera,
Olearia, Hedycarya and Bedfordia spp. invade
Myrtle Wilt gaps, burned areas and landslips in Victorian rainforest.
Little is documented on the distribution of weed species in Victorian
rainforests apart from the overview of information by Carr et
al. (1992). There have been no detailed studies of the relationship
between their distribution and disturbance regimes in Victoria
apart from that of Peel and Coram (1993) who found strong associations
between disturbance and weed establishment in the warm temperate
rainforests of the Lower Snowy River. Neyland and Brown (1994)
found a strong positive relationship between degree of disturbance
to Tasmanian cool temperate rainforest and the cover of weed and
pioneer species. The effect depended on the proximity of populations
of the weed species and their dispersal biology.
Wales (1972) and Ranney et al. (1981) showed that major
vegetational change in slow-growing northern temperate forests,
caused by microclimatic changes, extend 10-30m inside forest boundaries.
Laurance (1991) measured the impacts of agricultural clearing
on tropical rainforest in north east Queensland by the relative
abundances of disturbance-adapted plants, and found evidence of
disturbance, reflected in the abundance of such species, up to
500m inside the margins of rainforest fragments, although most
striking evidence was apparent within 200m of edges. He concluded
that isolated reserves must exceed 2000-4000 ha depending on shape,
to ensure that 50% of the reserve remains unaffected. The evidence
cited above suggests that increased light and soil temperature
and reduced humidity are likely to favour early successional and
sclerophyll plant species. Furthermore, site preparation burns
and fuel reduction burns tend to discourage both the more mesic
species such as Hedycarya and Olearia argophylla
and the advancing rainforest species such as Acmena smithii
and Pittosporum undulatum (Mueck and Peacock 1992).
Repeated intensive harvesting of sclerophyll forest is likely
to have some impact on the floristic composition of sclerophyll
forest itself. There are significant differences in the floristic
composition of the understorey of undisturbed old-growth forest,
naturally burned areas, and areas that have been logged in Victorian
forests (Griffiths and Muir 1991, Mueck and Peacock 1992, Ough
and Ross 1992). While the long-term outcome of these changes
in still uncertain, any changes that do occur may in turn be reflected
in the interactions between sclerophyll forest and rainforest.
One outcome may be increased abundances of opportunist and disturbance
specialists in the ecotone, increasing the potential for
invasion of mature rainforest by environmental weeds. Such vegetational
effects are likely to be difficult to detect without long-term
monitoring of the ecotone between rainforest and clear fall areas.
For the management of that portion of Victoria's rainforest that
may be subject to adjacent timber utilisation, the question is,
how adequate are forest buffers 20m to 40m wide in protecting
the ecological processes within rainforest? That is, how far
do the changed processes within clear fall areas penetrate into
adjacent forest, how long do these changes last, and what are
the implications of these changes? We must also ask, how closely
do these changes mimic the natural changes that in the absence
of harvesting would follow wildfire, in terms of kind, extent,
severity and frequency? These are important considerations for
the rainforest stands where the Code of Forest Practices applies.
Edge effects were defined originally by Leopold (1933) to represent
the increased number of species encountered where two major habitat
types intergrade. More recently, the term has been extended to
encompass the rapid creation of abrupt edges in extensive, relatively
undisturbed habitat, and the ecological changes that result from
such changes (Lovejoy et al. 1986, Reese and Ratti 1988).
Thus, an edge may be viewed as a marginal zone of altered microclimatic
and ecological conditions that contrasts with the forest interior
(Matlack 1993). For the purposes of this review, edge effects
refer to all measurable changes at an ecosystem boundary and within
adjacent ecosystems following anthropogenic disturbance. Changes
in the ecosystem usually are beyond the visual edge caused by
the impact. The ecological edge, or ecotone, that results from
a disturbance is the result of interactions between the kind and
intensity of the disturbance event and the ecological dynamics
within the adjacent, undisturbed environment.
CFL (1987) observes that the main threats to rainforest resulting
from activities in adjacent areas are increased risk of fire entering
rainforest, and increased exposure to wind and light following
removal of surrounding eucalypts and all other vegetation following
clear falling. Both of these assertions were made without reference
to any supporting evidence. There is little doubt that clear
felling of tall eucalypt forests causes important changes in the
ecological processes in adjacent forests: parameters of disturbance
such as gap size and distance from an abrupt edge are known from
many silvicultural and ecological studies to be closely associated
with changes in temperature, moisture, and associated fire ignition
probabilities (eg., Uhl and Kauffman 1990), precipitation, frost,
wind and fire behaviour (eg. Roberts 1973), insect activity (eg.,
Simandl 1992), nutrients (eg., Yanai 1991), light, and the composition
of soil borne bacterial and fungal populations (eg., Jha et
al. 1992, see the review by Bradshaw 1992). There are numerous
parameters which may affect rainforest across the boundary from
clear fall areas such as diseases, weeds, predators, soil destabilisation
and so on. These factors are reviewed below.
Exposure to windthrow
Potential impacts of the edge effects of harvesting activities
on rainforest fall into several main categories. Perhaps the
best understood of them is the effect on tree survival and vigour
of sudden exposure following clear fall operations, largely because
it has long been of interest for the design of good silvicultural
practices.
The structure of a forest affects the behaviour of wind in and
around it. Wind in turn affects water use, evaporative stress,
surface drying, disease spread, spray applications and wildfire
behaviour (Miller et al. 1991). For example, the likelihood
of windthrow can be influenced greatly by cutting practice, and
forest fires usually behave in direct response to local winds
(Reifsnyder 1955, Roberts 1973). Elevated tree mortality along
recently cut forest boundaries is a well established phenomenon
(see the review by Mayer 1989). There is little doubt that windthrow
occurs in Victorian forests, and such effects are not restricted
to the forest edge. Changes in vertical and horizontal wind velocities
and behaviour that result from clear felling can be detected several
hundred metres inside a forest, depending on things such as lapse
or inversion conditions, the extent of the edge perpendicular
to the wind, stand density and the density of the understorey
(Reifsnyder 1955, Rayner 1971, Miller et al. 1991). For
example, vertical wind profiles from a small, isolated stand of
ponderosa pine suggested that wind profiles in an infinite stand
are reached 75m from the forest edge (Reifsnyder 1955). In a
North American coniferous forest adjacent to a contiguous field,
wind speeds penetrating the forest edge in the trunk space were
significantly greater than those in the canopy for a distance
of about 60m, but with a longer fetch into the forest, wind speeds
varied little with height to mid canopy (Rayner 1971). Laurance
(1991) detected 'severe' canopy damage from wind in 36% of sites
within 150m of the edge in permanently isolated Queensland rainforest
stands. Windthrow was primarily responsible for tree mortality
in isolated stands of old-growth conifer forest with shallow roots
and loose soils. Stands up to 1 ha were entirely 'edge habitat',
experiencing 30% mortality (Esseen 1994). Edge orientation may
also be important in determining the effects of wind on tree mortality.
Its importance depends on exposure, determined by the location
of the edge in the landscape (Alexander 1964, Rayner 1971, Swanson
et al. 1988, Palik and Murphy 1990). Interpretation of
the quantitative aspects of these studies needs to take into account
the fact that linear rainforest stands mostly occupy watercourses
low in the landscape. However, rainforest buffers in production
areas are generally exposed, at one time or another, on all sides
as harvesting proceeds. For example, clear falling up or downstream
may increase wind speeds for airflows that follow drainage lines.
The concern for rainforest protection is that there will be windthrow
of eucalypts and rainforest species in the buffer, due to increased
wind velocities and to turbulence caused by the uneven canopy
(Morrison and Raphael 1993). If for no other reason, tree falls
on the buffer edge are likely to affect core rainforest because
eucalypts in rainforest buffers are considerably taller than 40m.
Creation of an abrupt edge may make trees in the buffer zone
susceptible to localised crown dieback. Because windthrow is
likely to affect forest for distances greater than 40m, it is
likely to affect the ecological dynamics of both the buffer community
and core rainforest, increasing fire risk, exposing sites to increased
light and reduced humidity, and predisposing sites to disease
infection. These effects are likely to be strongest in relatively
exposed rainforest stands with 20m buffers. There are no data
available on windthrow in Victoria's forests.
Microclimatic impacts
Clearfelling produces a sharp edge resulting in a loss of uniform
or gradually changing shade and moisture conditions from sclerophyll
forest to rainforest. The environmental gradients between the
two communities are sharply contracted. As a result, increased
light and soil temperatures, and reduced moisture levels may penetrate
the rainforest from the edge, resulting in a shift in the location
and a sharpening of the gradient. Disturbances across the boundary
of a rainforest may result in impacts on extant individuals, but
they may also affect the regeneration niche of rainforest species,
so the demographic characteristics of species may result in time-lagged
response to disturbance. This inference is based on the observation
that the recruitment dynamics of competing temperate rainforest
and sclerophyll species are governed by light availability, particularly
in zones of competitive overlap in ecotones. For example, N.
cunninghamii has a higher light compensation point than many
other rainforest species, but lower than E. regnans (Howard
1973b). Gleadow and Rowan (1982) found that the successful recruitment
of Pittosporum undulatum seedlings in a eucalypt forest
depends on microclimatic conditions, especially light intensity
and temperature, which ameliorate the effects of summer drought
on first year seedlings. Rainforest flora dependent on constant
moisture regimes (such as vascular and non-vascular epiphytes)
and shade dependent herbaceous species are most likely to be deleteriously
affected. However, there are no data on the responses of these
species to changes in light, temperature and moisture.
Some information is available on light and temperature changes
in Victorian ash forests following clear falling, gap selection
(0.25 ha) and shelterwood harvesting (50% retained BA). The following
information is based on a report by Saveneh and King (1994).
For mean minimum temperatures, partially harvested forests (gap
selection and shelterwood) were warmer than clear fall (by about
1.1oC), and for mean maximum temperatures, they were cooler than
clear fall (by about 0.7oC). Partially harvested forests experienced
higher minimum and lower maximum surface soil temperatures and
fewer frosts than clear fall areas. Shelterwood and unharvested
treatments received respectively 74% and 39% of the total global
radiation received by clear fall treatments. While these data
do not address directly the changes one might expect in rainforest,
or rainforest buffers, given clear falling in adjacent forest,
they do provide an indication of the magnitude of changes in these
parameters that one may expect. They also show that removal of
50% of the eucalypt basal area results in significant amelioration
of the changes in these physical parameters compared to clear
fall. One could reasonably expect a significant impact of landscape
position on micro-climate. In particular, air and soil temperatures
are likely to be lower in gullies, especially on sheltered sites
where solar radiation is reduced (S. Murphy, pers. comm.).
Light interception is closely related to canopy structure (Lowman
1986), but removal of adjacent eucalypt forest results in increased
light penetration to the understorey, even if there is no direct
overhead canopy disturbance (Crome et al 1992, Turton and
Duff 1992). The amount of the increase depends on the details
of the forest structure, topographic location, the height of the
forest, the structure of understorey vegetation, and the vegetation
structure in the edge periphery (eg., Lawton 1990). In the ecotone
between rainforest and sclerophyll forest, increased light levels
will favour those sclerophyll species that do not require mineral
soil or an ash bed for regeneration. These species may include,
for example, Acacia melanoxylon and Olearia argophylla.
Increased light levels usually attenuate quite rapidly from an
exposed edge, falling within a few percent of continuous forest
levels, similar to those experienced within small tree-fall gaps
in mature rainforest, within about 10m of the edge (Turton and
Duff 1992, MacDougall and Kellman 1992). Thus, a buffer of 20m
is probably sufficient to protect core rainforest from elevated
light, but it will still affect competitive dynamics within the
buffer, at least until adjacent sclerophyll forest regrows.
Other microclimatic variables attenuate less rapidly with distance
from a sharp boundary. There is abundant information of the effects
of clear falling on temperature regimes, and the interaction of
temperature and moisture. In general, removal of vegetation results
in very significantly increased temperatures in the cleared areas,
and increased temperature and reduced relative humidity in adjacent
forest that diminishes with distance from the edge. These factors
interact with light availability and wind to directly influence
tree survival (Whitehouse 1991) and are likely to affect associated
species, particularly relatively sensitive species such as epiphytes.
There are many examples of studies on changes in physical variables
following creation of a sharp forest edge. There are no such
data available for Victorian forests, but the changes depend only
on forest structure and physiognomy, so it is worthwhile reporting
the results of studies from a range of different forest types
from tropical to temperate ecosystems. Kapos (1989) and Williams-Linera
(1990) found that significant changes in the water status of isolated
Amazonian rainforest stands were detectable more than 20m from
the forest edge. Wind flow and increased insolation caused changes
of up to 20% in relative humidity at the edge of isolated Amazonian
rainforest stands compared to the interior and the whole of stands
smaller than 10 ha were affected by changes in relative humidity
and temperature, resulting in elevated tree mortality subsequent
to isolation (Lovejoy et al. 1986). Matlack (1993) detected
significant edge effects in a temperate, deciduous oak-chestnut
forest in light, temperature, litter moisture, vapour pressure
deficit, humidity and understorey cover up to 50m from the forest
edge. Chen et al. (1992) found physical and ecological
changes between 16m and 137m from a sharp edge in Douglas fir
forest, depending on the variable. These results are relevant
to Victorian rainforests because the physical effects of exposure
will be qualitatively the same. Changes in light, temperature
and relative humidity in adjacent stands will follow clear falling
operations. The consistency of the results reported from a range
of forest types suggests that we should expect the same kinds
of processes to operate in Victorian forests. There will be marked
variation in the responses depending on topography, forest type
and climate.
These changes are likely to affect the competitive ability of
different species differently, and may result in changes in the
composition of the flora of the protected stands of Victorian
rainforests. Desiccation is a key issue for moisture dependent
bryophytes (dependent on atmospheric moisture), epiphytic filmy
ferns, fork ferns (some of which are rare nationally and vulnerable
in Victoria) and orchids, some ground ferns such as Allantodia
australe and the rare Deparia petersonii, and some
herbs such as Australina muelleri ssp. muelleri.
Read and Busby (1990) found that several endemic Tasmanian rainforest
species are intolerant of high summer temperatures and low precipitation.
Hill et al. (1988) found significant differences among
temperate rainforest tree species in relative rates of photosynthesis
at different temperatures. Warm temperate rainforest species
Tristaniopsis laurina and Acmena smithii differ
significantly in their abilities to resist drought and desiccation
(Melick 1990b). Many Tasmanian rainforest dominants including
Lagarostrobus franklinii (Gibson and Brown 1991) and A. selaginoides
(Cullen 1987) rely on gap formation and periodic disturbance
resulting in greater light penetration for recruitment.
Studies of the competitive differences among rainforest taxa do
not emphasise the differences between rainforest and sclerophyll
taxa. Tropical (Amazonian) and northern temperate examples fail
to highlight the natural variability characteristic of the Australian
environment. These results suggest that changes in light, moisture
availability and temperature will favour some sclerophyll species
over some rainforest species, particularly near forest edges.
Their effects are likely to ameliorate over successional time.
As adjacent forest recovers, the transition zone will become
less abrupt and the effects will penetrate the buffer to a lesser
extent (Figure 4). If all coupes adjacent to a rainforest are
harvested within a short period of time, impacts will be more
severe over a shorter period. If coupes are harvested at longer
intervals, impacts will be less severe but more persistent. In
either case, rotation periods of 60-120 years are much shorter
than the recovery time of mature rainforest following crown fire,
which is probably greater than 200-300 years. This observation
does not suggest that clear falling of adjacent forest has impacts
equivalent to crown fire. Rather, it is made to emphasise that,
over extended periods, both the intensity and the frequency of
disturbance events may be important determinants of the severity
of impact.
Figure 4. Mean dominant
height of E. regnans and E. delegatensis (Forest
Information Section, pers. comm.). The data are for Site Index
33, based on the average of 1939 mountain ash (E. regnans)
plots from the Victorian Central Highlands, measured in 1994 (Forest
Information Section, CNR, pers. comm.)
For changes in many physical parameters related to, for example,
wind, light, and moisture regimes in rainforest, the effects are
likely to last less than 20 years, the time required for regenerating
ash forests to exceed the height of rainforest (Fig. 4). The
rate of this change will depend on the adjacent forest type and
its rate of growth, dependent to some extent on the silvicultural
treatment applied to the coupe (Bradshaw 1992). For example,
selective logging has the effect of increasing temperatures and
reducing relative humidity to a lesser extent than clear fall
harvesting (eg., Uhl and Kauffman 1990). The interpretation of
these results is complicated by the interaction of these factors
with windthrow and the incidence of fire.
Nevertheless, it is likely that even in the absence of interactions,
physical changes on the rainforest boundary after creation of
an edge will change the distribution and abundance of species
in the ecotone and probably within the rainforest itself. For
example, Barik (1992) found that significant changes in light
and moisture regimes along a gap size gradient play an important
role in influencing the composition and abundance of shade tolerant
and shade-intolerant tree species in gaps and affect the overall
species diversity of the forest in Indian subtropical broadleaved
forests. Altered vegetation structure and microclimatic conditions
are associated with the loss of epiphytic lichens, altered composition
of the understorey, increased mortality of perennial herbs, and
colonisation by disturbance specialists in boreal forests (see
Esseen 1994). It is yet to be seen if changes in physical parameters
will result in fundamental changes in the dynamics of the rainforest,
or will affect the long term persistence of rainforest stands.
It may be that the most significant risk of loss of rainforest
stands results from changes in physical parameters in surrounding
forest, leading to changes in fire dynamics. It may also
be that because rainforest has eveolved in a fire-prone environment,
rainforest species are adapted to the kinds and intensity of changes
in physical parameters that result from clear-falling. Estimation
of the extent of the impact of light, temperature, moisture and
windthrow effects in Victorian forests is, by necessity, speculative,
because no data are available. The application of adaptive management
principles would therefor suggest that until such data are available,
it would be best to assume that such affects are likely to have
impacts that are as important as those identified in other forest
ecosystems.
Impacts on fauna
The vast majority of studies of the effects of abrupt edges on
forest fauna have been on bird populations. Typically, studies
in temperate ecosystems report higher densities and diversities
of birds along edges than in community interiors (eg., Gates and
Gysel 1978, Strelke and Dickson 1980, Kroodsma 1984), a phenomenon
that falls under the general heading of an edge effect (in the
sense of Leopold 1933), and reduced populations of species dependent
on the conditions of the forest interior (eg., Ambuel and Temple
1983, Noss 1991). Changes in bird faunal composition depend on
the kind and severity of disturbance (DeGraaf 1992), stand size
and shape, feral predators, and the ecology of the species involved
(eg., for Amazonian forest fragments, Lovejoy et al. 1986,
Johns 1991). Edge-related increases to predation in nesting birds
may extend 150m-600m inside a forest edge in some kinds of forests
(Gates and Gysel 1978, Wilcove et al. 1986).
Such effects are relatively immediate and longer term, secondary
effects are possible associated with the disruption of ecological
process such as reproduction in long lived species and the disturbance
of pollinators or other factors that impinge on the reproductive
or regeneration niche of species (Lovejoy et al. 1986).
While the responses of birds may not be representative of the
majority of forest fauna, in the relatively few studies of edge
effects on taxa other than birds, impacts are qualitatively the
same. Certainly, other animals could be expected to respond to
changes in the physical environment near an abrupt edge. For
example, Lunney et al. (1991) found that light regimes
partly explain the distribution and abundance of three lizard
species in sclerophyll forest in southern NSW. Light and moisture
changes in Panamanian rainforest after logging in adjacent forest
are detrimental to the eggs and young of iguanid lizards, but
have no effect on adult survivorship, suggesting that the effects
of edge changes may be time-lagged (Williams-Linera 1990).
Victorian rainforest exists in patches. The theory of island
biogeography (MacArthur and Wilson 1967) and more recent theory
on metapopulation dynamics suggest that such archipelagos of habitat
will exchange individuals, and that this exchange will maintain
populations throughout the archipelago, even though species may
occasionally become extinct within individual stands. Such dynamics
will enhance species richness, and movement of individuals will
increase effective population sizes, thereby enhancing genetic
diversity within species. Wind, water, insect, mammal and bird
dispersal of propagules (either gametes or zygotes) are likely
to be involved in these processes. Loss of small stands, or interruption
of dispersal mechanisms by other means, will inhibit migration,
with detrimental effects on species richness and genetic diversity.
For example, many species that inhabit rainforest may have dynamic
interactions with adjacent seral and transitional communities,
and even with adjacent sclerophyll forest. Many rainforest species
have seeds with short viability (many less than 6 weeks, Floyd
1977, Hopkins 1990). Absences of mature eucalypts in adjacent
forest may affect pollinators and dispersal agents (insects, birds
and mammals). The proximity of seed sources, dispersal agents
and pollinators should be considered in estimating the impacts
of harvesting on rainforest. The ecology of plants and animals
inhabiting transition zones and sclerophyll forest probably play
an important part in maintaining ecological systems in rainforest
stands.
The most important conclusion, from the perspective of the protection
of Victorian rainforest, is that impacts of changes at forest
edges on animal populations are often species specific. This
makes generalisations concerning the impacts of edges on animals
difficult, and probably irrelevant because there is no information
currently available from Victorian ecosystems on the impact of
edge effects on individual rainforest animal species.
The size and shape of rainforest stands
An important feature of all rainforest communities is that they
occur in relatively small, more or less isolated stands, or in
linear, branched stands following drainage lines (Cameron 1992).
Rainforest is surrounded by other, usually sclerophyllous vegetation.
In the Central Highlands, for example, some cool temperate rainforest
stands follow water courses adjacent to damp forest, though most
are associated with wet forest and montane forest. Figure 5 shows
that the perimeters of cool temperate rainforest stands are much
longer than for a circle of the same area, reflecting their linear
(or dendritic) shape. This feature is qualitatively the same
for the other rainforest types and it has important consequences
for the design of protection measures.
Figure 5. The area / perimeter
ratios of cool temperate rainforest stands in Victoria. The dimensions
for warm temperate rainforest are qualitatively the same. The
upper envelope is the ratio expected for a circle of the same
area as the stand.
Area/perimeter ratios of small rainforest stands are closer to
the ratio for a circle of the same area than are the ratios of
large stands (Fig. 5). This phenomenon may be at least partly
a function of the fact that the data are based on aerial photo
interpretation at the scale of about 1:25 000 or 1:40 000, mapped
at 1:100 000. Stands of 1 ha were 1 mm2 when mapped, approaching
the size of the width of the pen used for mapping, and the details
of dendritic shapes cannot be mapped at that scale. Figure 6
shows that most Victorian rainforest stands are small, less than
2 ha. Using CNR's definition of rainforest, the largest stand
of rainforest recorded on the Departmental database is a stand
of warm temperate rainforest in East Gippsland, totalling 230
ha. The largest stand of cool temperate rainforest is 189 ha,
and the largest stand of overlap between cool and warm rainforest
is 110 ha (but see Fig. 6b).
Figure 6a. Frequency distribution
of cool temperate rainforest stand size.
[
Figure 6b. Frequency distribution
of warm temperate rainforest stand size. The two patches greater
than 200 ha are mostly overlap communities between cool and warm
temperate rainforest (CNR Forest Information Section, pers. comm.)
Extent of edge effects is, in general, a function of the variable
under consideration, the length of the edge relative to the size
of the undisturbed stand, the ecological conditions in the surrounding
habitat, and the degree of contrast between different stand types
(Angelstam 1986, Reese and Ratti 1988, Laurance 1991, Franklin
1993). Edge width is critical because smaller stands have a relatively
high edge/interior ratio (eg., Diamond 1975, Franklin 1993).
Processes such as radiation, wind exposure, wind behaviour, temperature
and drying interact with stand shape, aspect and topography resulting
in site-specific degrees of effect (Matlack 1993). But in general, |
